Innovative solutions to long-standing challenges are being provided by artificial intelligence, which is revolutionizing healthcare at a breakneck tempo. AI has emerged as a promising instrument to address the gaps in specialist shortages and the increasing costs of healthcare in South Korea, where an aging population has placed a strain on the medical system. However, as AI becomes more integrated into medical decision-making, experts are questioning whether it should continue to serve as an assistant or assume a more autonomous role.
Germany and Sweden are the two countries that are at the forefront of this unprecedented transformation in the application of AI in diagnosis. Sweden has implemented artificial intelligence (AI) into its mammography double-reading systems. Germany has mandated the use of AI as the primary diagnostic instrument for the majority of conditions.
This global momentum is incapable of being disregarded. For instance, Coreline Soft, a Korean medical AI company, concurs that AI in health care will continue to expand, even if Korea does not intend to implement such a policy in the near future.
Prof. Choi Joon-il of Seoul St. Mary’s Hospital anticipates that AI will substantially diminish the necessity of physicians in the medical field in the near future. “AI will be present in nearly every medical field within the next decade.” He further explains that in the next two decades, it will assume responsibility for the majority of the supportive responsibilities that physicians presently manage.
Nevertheless, not all of this is advantageous in this context. Jeong Won-gi lectures at Chonnam National University Hwasun Hospital, where he was engaged in the development of AI tools for the diagnosis of pulmonary nodules. He observes, “AI has indeed simplified our workflow; however, it is still incomplete.” It occasionally misidentifies a non-nodule as a nodule, resulting in unnecessary anxiety.
According to a study published in the European Journal of Cancer, AI outperformed human physicians by a margin of 31 errors in the detection of large pulmonary nodules, while the doctors made 407 errors. Despite the significant progress made, the infrequent errors made by AI necessitated human supervision.
The function of artificial intelligence in emergency medicine
Professor Kim Jin-young of Soon Chun Hyang University Hospital in Cheonan believes that he would query the application of AI in care. He was employing VUNO Med-DeepCARS, an AI system that enabled the prediction of patient cardiac arrest cases. He asserts that this has its advantages; however, the human element remains absent. “Our system is limited in its ability to efficiently process data.” The nurses are required to perform the task manually on the vital sign side. False alarms are prevalent. He further stated that the hospital personnel would be inundated with work in no time.”
Kim’s concern is considerably more profound. Artificial intelligence is devoid of human discernment.It is incapable of comprehending the intricacies of patient care, emotions, or context. That is an attribute that technology will never be able to replicate.
Human Touch and Patient Trust
However, the ultimate objective remains patient trust as the day concludes: “AI is capable of rapidly interpreting chest X-rays, which is invaluable in emergency situations.” However, patients ultimately rely on their physicians, not a machine, as stated by Lee Hae-min, President of Incheon Christian Hospital.
Lee contemplates the potential for AI to integrate medical history, symptoms, and imaging data into a comprehensive diagnosis. However, he continues to perceive the sentiment that physicians should maintain the authority to make decisions at all times. “No matter how advanced AI might be, it is the doctors alone who can provide empathy, reassurance, and accountability.”
In this context, Professor Choi cautions that the loss of expertise should not be the consequence of AI: “In the past, we diagnosed patients using stethoscopes and instinct.” In the present day, numerous physicians are unable to function without a CT scan. The next generation of physicians may not be aware of when the system is making an error if they exclusively rely on AI.
In spite of this, the field of medicine will experience a growth in the applications of AI. It has been demonstrated to reduce diagnostic time and efficacy. According to a study published in The European Journal of Radiology, the chest X-ray triage time was reduced by 77% by Lunit INSIGHT CXR.
AI is a tool, not a substitute, as the specialists have stated. Lee concludes, “AI is present to assist us, not to compete with us.” “It enhances what we do, but it will never replace the human connection that defines medicine.”
The challenge will be to strike a balance between human expertise and technological advancement in the pursuit of AI to influence the future of healthcare.
